An Integrated Population Pharmacokinetic and Machine Learning Model for Predicting Tacrolimus Exposure in Adult Patients with Nephrotic Syndrome - PubMed
6 hours ago
- #Machine Learning
- #Tacrolimus
- #Pharmacokinetics
- The study developed an integrated model combining population pharmacokinetics (PPK) and machine learning (ML) to predict tacrolimus (TAC) exposure in adult nephrotic syndrome patients.
- Incorporating individual PK parameters as features improved predictive performance (R²: 0.633 vs. 0.602) compared to models without PK parameters.
- A weighted ensemble of CatBoost, AdaBoost, and GraBoost (5:3:2) achieved the best performance with R² = 0.633, MAE = 1.081, and RMSE = 1.377.
- SHAP analysis identified PK parameters as the most influential feature, surpassing conventional biochemical indicators.
- The integrated approach supports individualized TAC dosing, enhancing safety and precision in clinical applications.